The advent of the optimization machine learning mechanisms can hardly be underestimated – during 2021-2022 around 47% of business owners reported that thanks to ML and AL they were able to optimize sales and marketing, and 32% reported that they were able to reduce operating costs.

Smart algorithms in ad tech have a paramount role since it is a core of all processes that increase bidding effectiveness, improve ad delivery, and maximize the income you receive as a marketplace owner. 

Over the course of this article I want to talk about how setting the important traffic optimization parameters will help your adtech system to operate automatically and more efficiently.

Let`s start with a QPS adapter . This is a leverage thanks to which you can effectively match traffic from supply partners with demand partners that need it most. Such demand partners also normally offer the best bidding rates for inventory so it means that inventory will be sold for the most competitive price.

As bidding will turn more successful it means that more traffic will be purchased in your marketplace so it will generate more profits. This feature constantly analyzes the incoming traffic to make it work more productively for the whole ecosystem. In other words, it determines what kind of traffic is most frequently bidded on by each DSP, and based on this insight the system distributes the traffic in the future.

During this process a variety of parameters can be taken into account, for example: 

Most parts of the traffic that goes through the adapter will be sorted this way. A little fraction of the remaining traffic is also passed to DSP so that the system could learn the new parameters and adapt to changes. A QPS adapter works on the demand side so you need to activate the QPS adapter checkbox when you create a new connection on the demand side.

A traffic logger is also a very important optimization tool that helps you to review how much and what kind of traffic your marketplace receives in detail. 

How does it become possible? With the traffic logger turned on, the system analyzes the traffic that your programmatic platform obtains. Also, with the help of the traffic logger, it is possible to define the bidfloors, geo, and tmax of your supply and whitelist for the DSP partner only those SSPs that are sending suitable traffic. This way you will be able to review the statistics and understand the ‘anatomy’ of your traffic in detail. 

Driving more highly converting and relevant traffic to your marketplace is the right thing, however, it is also important to eliminate undesirable traffic that doesn’t correspond to your quality criteria. 

For example, you can exclude certain sources by indicating their parameters in the blacklist section. Blacklists with extensive filters also help a lot. Website domains, applications, bundles – these all can serve as important attributes according to which you sift the traffic through. You can block the existing endpoints worldwide. 

“The quality of the traffic, the performance of your integrated partners, and the distribution of the traffic, these aspects are all in your hands to manage.”

On the contrary, you can prioritize the traffic sources by applying whitelists. In this case, you will select the criteria according to which your system will approve the traffic. 

Smart algorithms can also work on the background of many processes. The adtech algorithms which are so-called ‘quality control’, guarantee you a traffic quality. When the requests from SSPs don’t meet the desired criteria, then they are dropped. At the same time, when this issue becomes common, it is important to understand the cause of this problem. 

RCPM stands for real CPM (cost per mille, or cost per thousand impressions), and it is extremely useful for both demand and supply partners. Normally, it shows publishers the true value they are getting from an ad partner. This value is calculated based on the real number of impressions that ad units are getting on the publisher’s inventory, (not just impressions the ad partner agreed to pay for).

The calculating formula of DSP RCPM is DSP Spend/Bid Requests *1000000. This metric shows the performance of requests on the DSP side depending on spend helps analyze which traffic is more profitable.

A lot of companies that work in media trading (RTB) think that there is a direct correlation between their QPS plan and the profits since ad formats, traffic types, and optimization tools applied will also significantly impact the outcomes of the media trading. But this is a wrong statement.

However, the quality of the traffic, the performance of your integrated partners, and the distribution of the traffic, these aspects are all in your hands to manage. Using the leverages described in this article, you can create with a high possibility, a beneficial media trading, configure it, analyze and set it up your platform for big success.